Lead Scoring: Five Steps to Getting Started How-To Guide

How-To Guide
© 2013 Demand Metric Research Corporation. All Rights Reserved.
Getting Started with Lead
Scoring
By David Raab, CEO at Raab Associates
September 30, 2013
EXECUTIVE SUMMARY
Lead scoring applies mathematical formulas to rank potential customers. It
is chiefly used to identify prospects that are ready for direct sales contact.
Because the calculations are automatic, the scores are consistent, current,
and can include more variables than any manual assessment. This saves
marketers work, ensures that all qualified leads are sent to sales promptly,
and keeps non-qualified leads out of the sales system.
Despite the obvious value of automated scoring, a recent survey by
Aberdeen Group found it’s used by barely more than half (57%) of
marketing automation users. The remainder may not know how to set up a
lead scoring program or may consider it too much work. This paper will
help those marketers to get started by showing how to set up a simple
system and then refine it over time.
THE CASE FOR LEAD SCORING
Before investing in a lead scoring system, most marketers will want a clear
picture its benefits. The main use for lead scoring is selecting leads to send
to sales. Doing this with an objective scoring process will screen out
unqualified leads and find qualified leads that might otherwise be
overlooked. Consistent quality encourages sales people to invest their own
time in pursuing the leads they receive. Most lead scoring projects also
define formal processes for lead transfer, set standards for how quickly and

How-To Guide
how many times sales people will attempt to reach marketing-qualified
leads, and return rejected leads to marketing for additional nurturing.
These changes ensure greater utilization of marketing-generated leads and
less need for sales people to prospect on their own.
But there’s more to lead scoring than deciding when to transfer leads.
Scores provide an objective measure of lead quality, helping marketers
understand the true value of different acquisition programs, spot trends in
lead quality, and identify changes in behavior. Sales people can use scores
to prioritize their calls and to better understand individual leads. Scores
can identify each prospect’s stage in the buying cycle, helping to track
movement between stages, select marketing treatments, and use stage
counts to build better sales forecasts.
STEP BY STEP
Ironically, the first step on your lead scoring journey is to forget most of
the benefits listed above. Instead, focus on a single, simple goal: using lead
scores to select leads to hand over to sales. This is enough to get started.
Once you have that application running smoothly, you can think about
adding others. Once your attention suitably focused, you’re ready to dig
into specifics. Here are five steps for building a basic scoring system:
1. Set up the team. From marketing, you’ll need people who
understand existing acquisition programs and operation of the
marketing automation system. From sales, you’ll need sales reps
who are recognized experts in identifying quality leads, both so
they can give good advice and so their participation gives
credibility to the results. You’ll also need someone from sales
operations who can extract CRM data for analysis and can help
define hand-off processes. You’ll want senior managers from sales
and marketing who can agree on reasonable standards for
contacting new leads and for measuring results.
2. Identify the leads to select. Are you trying to find leads that will
eventually become customers or are you trying to find leads that
sales will accept and pursue? The groups are not necessarily
identical: sales might like leads from big companies where a deal
would be large even if the probability of success is low; or they
may prefer small deals that will close quickly; or they might
accept marginal leads physically near existing accounts. Chances
are it’s a mix of all those considerations and more.
CHECKLIST (1):
Set up the Team
 Sales Rep, Sales
Operations, Manager
 Marketing Operations,
Manager
 Project Manager
 Executive Sponsor
 Project Charter
 Project Plan with Dates

How-To Guide
Since the first goal is a scoring system that is accepted by sales,
you should start with scores that identify leads that sales will
want. Once the scoring system has more credibility and you have
more historical data available, you can try identifying additional
leads that are worth pursuing even though sales would have
ordinarily rejected them. But that comes later.
Fitness measures: The simplest way to identify leads that sales
will accept is to look at leads which sales has already accepted.
Pull a sample of accepted and rejected leads from the CRM
system and review them with sales to identify the characteristics
that led to sales’ decision. You’ll probably end up with a small
number of “demographic” traits such as industry, company size,
and personal title. Location is often on the list, especially when
in-person sales calls are common. Demographics might be
supplemented by traditional “BANT” qualifiers: budget (is this a
budgeted purchase?), authority (can this person make the
purchase decision?), need (does the company need your
product?), and timeline (when is the purchase likely to happen?).
But accurate BANT information is often not available until later
in the buying cycle.
Demographics and BANT are both measures of “fit”, which is
whether someone looks like a potential customer.
Engagement measures: Many salespeople have been trained to
immediately contact every high-fit lead they see. But a marketing
automation system will often capture many more leads than a
salesperson has been getting, including many at an earlier stage in
the purchase cycle. So even if the sales department doesn’t raise
the issue, you may want your scoring system to filter out leads
who are not ready for a sales contact.
This contact-readiness is generally called “engagement”. It is
measured by looking at behaviors captured in the marketing
automation system, including Web site visits, email responses,
Webinar participation, and downloads. Most of this information
has not previously been available to salespeople, which is another
reason they are unlikely to suggest including it in a scoring
formula. Marketers may wish to bring up some hypothetical
CHECKLIST (2):
Identify Leads to Select
 Define Goal: Acceptance
or Close?
 Review Sample of Past
Leads
 Choose Fitness Measure
 Set Engagement
Threshold
 Choose Engagement
Measures
 Estimate Lead Volume

How-To Guide
examples to drive home the value of screening on engagement:
“Do you really want to call someone who has only downloaded
one white paper and not opened four emails in the past month?”
Sales will probably agree that leads should reach some sort of
engagement threshold, such as a 10% likelihood of making an
appointment. You may not be able to calculate this at first, but
you can later track actual results and adjust the engagement
standards to match the target. If the data is available, supplement
your engagement threshold with an estimate of how many leads it
would require the salesperson to follow up.
3. Select the data elements to score. Once sales and marketing have
jointly decided what a qualified lead looks like in terms of fit and
engagement, they can begin to explore at the actual data available
for scoring.
For fitness measures, the key challenge is that the data must be
available in the marketing automation system. For example, the
sales may have found that only companies over $100 million
revenue are likely customers, but marketing automation may not
have captured revenue at the time of scoring. Faced with such
situations, the team may choose an alternative measure that is
already stored in marketing automation, change marketing
automation forms to gather the original item, append the
information from an external data source, or drop the element
altogether. Or, since no single approach is likely to provide data
on all leads, the company might apply several of these approaches
simultaneously.
For engagement measures, a great deal of data is usually available.
The challenge here is deciding which is relevant. Look at
whatever historical data is accessible, ideally examining the
behavior histories of individuals to identify typical patterns. You
might find that consuming certain pieces of content is highly
significant, such as a vendor comparison paper or pricing Web
page. If the existing data isn’t organized properly, you’ll probably
have to start with team members’ educated guesses. Organizing
the behavior data for scoring may require adding flags for events
such as visiting a particular Web page or maintaining summaries
such as the number of emails opened in the past two weeks.
CHECKLIST (3):
Select Elements to
Score
 Identify Relevant Data
Elements
 Assess Availability &
Accuracy
 Look for Predictive
Behavior Patterns
 Add Summaries & Flags
for Scoring
 Gather or Import New
Data as Needed

How-To Guide
As with fitness measures, there may be some external sources that
capture relevant behaviors, such as systems that monitor social
media. But you can probably defer use of these and other non-
marketing automation data sources during your initial scoring
project.
For both fitness and engagement measures, it’s important to look
beyond the list of available elements to assess the quality of the
actual data. Data is often missing or inaccurate, especially if it
isn’t used by the person or system who gathers it. Personal
contact information and BANT data in particular are often
unreliable. Marketers can work to improve data quality over time,
but in the short run, the scoring system will have to use what’s on
hand.
4. Define the scoring formula. Once you’ve assembled the list of
relevant data elements, you still must combine them in a formula
that produces the score itself. Ideally, this would be based on
statistical analysis that correlates the actual values with outcomes
for previous leads. But most companies don’t have the necessary
data or statisticians available. You can probably run some simple
reports, however. Go back to your sample of closed leads (see
Step 2) and compare the values of the selected data elements on
good vs. bad ones. Be sure to use the values as they stood when
the lead would have been scored, not at the end of the buying
process. This is important because demographic and BANT
values are often missing early in the buying cycle and behavior
measures accumulate over time.
If no data is available, you’ll have to rely on the educated
opinions of your team members. Either way, the result looks the
same: a formula with points assigned to different values of each
element. For example, you might assign ten points to companies
in your core customer industries, five points to companies in
peripheral industries, and zero points to industries that rarely use
your product. You might even assign negative points to industries
that clearly cannot use you product, to ensure those leads are not
passed on. Be sure to explicitly define treatment of missing data:
does it disqualify a lead, count as zero, or count as a middle
value?
CHECKLIST (4):
Define the Scoring
Formula
 Assign Points to
Different Data Values
 Consider Points for
Missing Values
 Compare Score to Sales
Rating on Sample
 Set Score Ranges &
Treatment Matrix
 Phone Calls to Classify
Marginal Leads

How-To Guide
The best answer will vary for different elements. Behavior scoring
works the same way, except that the input is often a flag or
summary field such as number of email clicks in the past week. In
general, keep things simple during the first scoring project; you
can always add complexity as you gather more data from actual
results.
Lead Treatment Matrix
Engagement
Fitness
You’ll probably want to develop separate scores for fitness and
engagement. Group these into ranges on a two-dimensional
matrix and specify which combinations are passed on to sales.
This gives more precise control than simply adding the two scores
together. For example, a high fit score with medium engagement
might get passed to sales, while medium fit and high engagement
might not, even though the sum of the two scores could be the
same in both cases. There might be other treatments as well, such
as sending marginal leads to a telephone qualification team and
sending very low quality leads to a generic monthly newsletter.
Once you’ve defined your initial formula, test it by having the
sales members of your team examine some sample leads and
decide how they would have rated them; then, score the leads
and compare the scores with the ratings. Explore any major
discrepancies – what did the sales people consider that the
scoring formula missed? Adjust the formula and try again until
the scores come reasonably close to the subjective ratings.

How-To Guide
When setting your cut-off points for which leads get passed to
sales, bear in mind that the cost of a missed opportunity is much
higher than the cost of contacting and rejecting an unqualified
lead. This means the standard for passing a lead to sales should
be fairly low. But you may want a higher standard during the
initial deployment of the scoring system, when it’s critical that
sales people see that the system is passing them high quality leads.
If you do start with high standards and lower them over time, be
sure that salespeople know what’s going on.
5. Deployment. Your scoring rules and boundaries need to be
transferred into the marketing automation system to become part
of regular operations. Test carefully to check that calculations
and selection processes accurately reflect the team’s intentions.
Have team members review the initial scored records to ensure
they meet the expected standards for quality and consistency. Set
up reporting systems to track process measures such as the
number of scored records, distribution of scores, and proportion
of missing data in each element. The marketing automation team
should review those measures regularly and explore any
significant deviations over time.
At the same time you’re finalizing the technical deployment, your
team will need to build operational processes to ensure the
selected leads are treated correctly. Necessary processes include
transfer of leads into CRM, assignment of leads to salespeople
(typically handled within CRM), and training of salespeople in
the agreed standards for initial contact and disposition reporting.
You might also ask salespeople to rate each lead as they receive it.
This provides immediate feedback on perceived quality and
identifies any problems with the scoring formulas. Be sure it’s
easy for salespeople to send these ratings and for marketing
automation staff to load them into the system. You’ll need other
reporting to ensure that the sales team is meeting the agreed-
upon treatment standards. Make sure everyone agrees that sales
will attempt to contact all leads they are sent, even if the
salesperson rates them poorly. Results from low-rated and low-
scoring leads are essential for fine-tuning the scoring formulas.
CHECKLIST (5):
Deployment
 Load Scoring to
Marketing Automation
 Agree on Transfer
Processes & Deploy
 Gather Sales Ratings on
Scored Leads
 Monitor System
Operations
 Compare Scores with
Actual Results
 Adjust Scoring Formulas
Based on Data

How-To Guide
First impressions are critical, so watch results carefully during the
initial period. Set up a formal monthly review of the leads sent,
having sales team members rate them and compare their ratings
with system scores. As before, explore the reasons for any major
discrepancies and fix any obvious major flaws.
By the end of the deployment period, the initial leads should be
far enough into the sales process that you can begin to see which
will become actual customers. Compare that information with
both the initial lead scores and salesperson ratings. This will help
to see whether the manual ratings are actually more accurate than
the scores: if not, there’s little reason to change the scoring
formula to mimic the ratings. As more results become available,
you can do other analysis to identify new data elements and
formula changes that would produce more accurate scores.
MEASUREMENT
Initial deployment and acceptance of your scoring system is just the start.
It’s critical to regularly review your scoring processes to ensure they are
operating correctly and to identify opportunities for improvement. Key
performance measures include:
 Scoring Operations: Look for trends in measures such as the
number of leads scored, number of leads sent to sales, and firing
rate for each scoring rule.
 Score Accuracy: Lead scores at the time of hand-off should
correlate closely with initial sales acceptance and final purchase
behavior. Although the final purchase is what matters most, it’s
often easier to capture accurate information on sales acceptance.
 Productivity: A better scoring system should ultimately result in
higher productivity for salespeople, as they rely more on marketing-
generated leads and less on their own prospecting efforts. Specific
metrics to track this include close rate, deal size, cycle time, revenue
per sales person, and percentage of revenue from marketing-sourced
leads.
 Results: Higher productivity should be reflected basic measures of
salesperson performance, including percentage of salespeople
meeting quota and percentage of revenue plan achieved.

How-To Guide
REFINEMENTS
Your initial lead scoring formulas will intentionally be kept simple to speed
deployment. But once the system is stable, you can consider refinements to
improve results. Some options include:
 Time-based Depreciation: Reduce the number of points earned for
older behaviors. This applies largely to engagement measures, which
should focus primarily on current behavior.
 Category Caps: Limit the total points available for any particular
type of behavior, such as visiting a Web page. This ensures the score
is not skewed by one repeated activity, such as downloading the
same paper several times because the visitor failed to save it.
 Negative Points for Disqualifying Activities: Assign very large
negative values to activities or attributes that make it clear someone
is not a potential buyer, such as visiting the Careers page or working
for a competitor. Alternately, you can do this with filters or
selection rules.
 High Points for Hot Leads: Assign very large positive values to
activities or attributes that should lead to immediate sales contact,
such as a “call me” request or download by high-priority target
account. Again, it might be better to handle this through campaign
rules rather than the scoring system.
 Different Rules by Customer Segment: Apply different scoring
rules to prospects in different customer segments, based on region,
role, company size, etc. This reflects differences in the definition of
a qualified lead within each group.
 Account-level Scoring: Increase scores when multiple prospects
from the same account are active at the same time. This might
indicate high interest in a purchase.

How-To Guide
EXPANSION
You’ll also want to expand use of the scoring system. Opportunities
include:
 Separate Scores for Different Products: These can be used to select
which product to offer new customers and to identify opportunities
for selling new products to current customers.
 Buying Stage: Scores can measure the prospect’s position in the
purchase cycle to help select content and other treatments. This
could be accomplished with a single score, by assigning a different
point range to each stage, or by creating separate scores for each
stage.
 System Decisions: Scores can be used in system rules to select
campaigns, guide treatments, assign segments, define contact limits,
alert salespeople, and do other tasks.
 Sales Insights: Scores can help salespeople to understand individual
leads. This requires showing the scores to the salesperson, usually
within the CRM system. It also needs an explanation of what
scoring levels mean and often includes access to attributes and
behaviors that contributed to a score.
BOTTOM LINE
Lead scoring can be complex, but marketers shouldn’t let that prevent
them taking advantage of it. They can start with a simple scoring system
focused on selecting leads that are ready for sales contact. The process
starts with a joint marketing/sales team, which will define the leads to be
selected, identify data elements to use in the scoring, build the scoring
formulas, and deploy the results. Keys to success include: a properly staffed
and motivated team; building scores that match salesperson ratings of new
leads; ensuring new leads are contacted by sales; and carefully measuring
results. Companies that follow this process will quickly gain immediate
benefits from lead scoring and have a solid foundation for future growth.

How-To Guide
ABOUT THE RESEARCH ANALYST
With an MBA from Harvard, David is an expert in both B2B &
B2C marketing strategy & technology. He has advised The Gap,
JC Penney, Lowe's, Blue Cross/Blue Shield, Williams-Sonoma,
Scholastic, Unisys, Sprint and Verizon Wireless.
He also publishes the Raab Guide to Demand Generation
Systems and the Marketing Performance Measurement Tool-Kit.

Lead Scoring: Five Steps to Getting Started How-To Guide

Recommandé

Recommandé

Contenu connexe

Tendances

Tendances (20)

En vedette

En vedette (6)

Similaire à Lead Scoring: Five Steps to Getting Started How-To Guide

Similaire à Lead Scoring: Five Steps to Getting Started How-To Guide (20)

Plus de Demand Metric

Plus de Demand Metric (19)

Dernier

Dernier (20)

Lead Scoring: Five Steps to Getting Started How-To Guide